…57304)

So far, `PartialStruct` has been unable to represent non-contiguously
defined fields, where e.g. a struct would have fields 1 and 3 defined
but not field 2. This PR extends it so that such information may be
represented with `PartialStruct`, extending the applicability of
optimizations e.g. introduced in #55297 by @aviatesk or #57222.

The semantics of `new` prevent the creation of a struct with
non-contiguously defined fields, therefore this change is mostly
relevant to model mutable structs whose fields may be previously set or
assumed to be defined after creation, or immutable structs whose
creation is opaque.

Notably, with this change we may now infer information about structs in
the following case:
```julia
mutable struct A; x; y; z; A() = new(); end

function f()
    mut = A()
   
    # some opaque call preventing optimizations
    # who knows, maybe `identity` will set fields from `mut` in a future world age!
    invokelatest(identity, mut)
   
    isdefined(mut, :z) && isdefined(mut, :x) || return
   
    isdefined(mut, :x) & isdefined(mut, :z) # this now infers as `true`
    isdefined(mut, :y) # this does not
end
```

whereas previously, only information gained successively with
`isdefined(mut, :x) && isdefined(mut, :y) && isdefined(mut, :z)` could
allow inference to model `mut` having its `z` field defined.

---------

Co-authored-by: Cédric Belmant <cedric.belmant@juliahub.com>
Co-authored-by: Shuhei Kadowaki <aviatesk@gmail.com>

…Lang#57222)

When a variable has a field set with `setfield!(var, field, value)`,
inference now assumes that this specific field is defined and may for
example constant-propagate `isdefined(var, field)` as `true`.
`PartialStruct`, the lattice element used to encode this information,
still has a few limitations in terms of what it may represent (it cannot
represent mutable structs with non-contiguously defined fields yet),
further work on extending it would increase the impact of this change.

Consider the following function:
```julia
julia> function f()
           a = A(1)
           setfield!(a, :y, 2)
           invokelatest(identity, a)
           isdefined(a, :y) && return 1.0
           a
       end
f (generic function with 1 method)
```

Here is before on `master`:
```julia
julia> @code_typed f()
CodeInfo(
1 ─ %1 = %new(Main.A, 1)::A
│          builtin Main.setfield!(%1, :y, 2)::Int64
│        dynamic builtin (Core._call_latest)(identity, %1)::Any
│   %4 =   builtin Main.isdefined(%1, :y)::Bool
└──      goto JuliaLang#3 if not %4
2 ─      return 1.0
3 ─      return %1
) => Union{Float64, A}
```

And after this PR:
```julia
julia> @code_typed f()
CodeInfo(
1 ─ %1 = %new(Main.A, 1)::A
│          builtin Main.setfield!(%1, :y, 2)::Int64
│        dynamic builtin (Core._call_latest)(identity, %1)::Any
└──      return 1.0
) => Float64
```

---------

Co-authored-by: Cédric Belmant <cedric.belmant@juliahub.com>

…uliaLang#57304)

So far, `PartialStruct` has been unable to represent non-contiguously
defined fields, where e.g. a struct would have fields 1 and 3 defined
but not field 2. This PR extends it so that such information may be
represented with `PartialStruct`, extending the applicability of
optimizations e.g. introduced in JuliaLang#55297 by @aviatesk or JuliaLang#57222.

The semantics of `new` prevent the creation of a struct with
non-contiguously defined fields, therefore this change is mostly
relevant to model mutable structs whose fields may be previously set or
assumed to be defined after creation, or immutable structs whose
creation is opaque.

Notably, with this change we may now infer information about structs in
the following case:
```julia
mutable struct A; x; y; z; A() = new(); end

function f()
    mut = A()
   
    # some opaque call preventing optimizations
    # who knows, maybe `identity` will set fields from `mut` in a future world age!
    invokelatest(identity, mut)
   
    isdefined(mut, :z) && isdefined(mut, :x) || return
   
    isdefined(mut, :x) & isdefined(mut, :z) # this now infers as `true`
    isdefined(mut, :y) # this does not
end
```

whereas previously, only information gained successively with
`isdefined(mut, :x) && isdefined(mut, :y) && isdefined(mut, :z)` could
allow inference to model `mut` having its `z` field defined.

---------

Co-authored-by: Cédric Belmant <cedric.belmant@juliahub.com>
Co-authored-by: Shuhei Kadowaki <aviatesk@gmail.com>